In an attempt to finally get a college diploma, I've started taking classes again. I signed up for a senior-level statistics class because I don't like calculus. This class however, turns out to concern calculus-based statistics--integrating over probability spaces, etc. Bummer.
I much prefer the statistics generated by catching people with fake ID's. Having set my 1991 goal at 1,000, I've caught 887 1/2 so far. Of those, about 20 percent have been false ID's obtained through various state departments of motor vehicles by people brazen enough to go to the DMV with fraudulent documents. These are interesting statistically because they contain a date of issue, so that I can tell how long they've been in use. It's hard to tell how long someone has been using a borrowed or altered ID, but it's easy to refer to the issue date on an official-but-fake ID. Given my lazy desperation to publish something in this issue of Noesis, I hereby present "months of use" statistics for a sample of 71 official-but-fake ID's caught from late May through early September of this year.
Numbers on the left represent the number of months an ID has been in circulation when I catch it. Dots represent the number of ID's in each category. Over half of the ID's of this type are returned to their users, so most of these remain in circulation.
A low percentage of this type of ID is detected and confiscated. Probably a much greater proportion are lost, stolen or accidentally destroyed. Most are probably simply outgrown as the user turns 21. Occasionally, someone who is of age loses their real ID (or is afraid to go to the DMV and get a real license) and continues to use their fake ID though they are of age. Somedmes we catch very old fakes used by people who are careless because they are about to turn 21.
My overall rate for misidentifying a real ID as fake is about 50 fake ID's accurately identified for every real ID I misidentify. Another error I could make is misidentifying a borrowed ID as an official-but-fake ID. I am confident that no more than three members of this sample of 71 are included because of rnisidentification errors.